| What is the best form of arguments from design?
post-template-default,single,single-post,postid-15815,single-format-standard,op-plugin,ajax_fade,page_not_loaded,,vertical_menu_enabled,side_area_uncovered_from_content,qode-theme-ver-10.0,wpb-js-composer js-comp-ver-4.12,vc_responsive

What is the best form of arguments from design?

What is the best form of arguments from design?


The goal of this essay is to find the best way to formulate an argument from design to the existence of God. In order to do that, I compare three epistemic approaches to the question “How do we detect design?”: the Reidian Perceptual Model (RPM), Inference to the Best Explanation (IBE) and Bayes’s theorem (BT). RPM turns out to be astonishingly robust but is vulnerable to objections from disagreement and fails to meet all criteria of sense perception. IBE is a more general, less formalized and ultimately weaker version of BT, wherefore only the latter should be chosen among the inferential models. A closer look at BT shows that it fills the gaps of RPM. Therefore, I present a fourth model which is a synthesis of RPM and BT and test it with the question “Do living beings exhibit divine design?”.

1.   Introduction

Because that which may be known of God is manifest in them; for God hath shewed it unto them. For the invisible things of him from the creation of the world are clearly seen, being understood by the things that are made, even his eternal power and Godhead; so that they are without excuse…[1]

So-called “Design arguments” are teleological arguments that seek to show from the existence of some contingent thing that God designed it and thus a fortiori God exists. They can be classified into Global Design Arguments (GDA) and Local Design Arguments (LDA)[2]. The former are roughly synonymous with Cosmological Design Arguments, most notably the Fine-Tuning Argument, whereas the latter deal mainly with living things (therefore sometimes called “Organismic Design Arguments”, ODA). For epistemic reasons, the following discussion is restricted to ODA.

Let me make a quick point concerning the relevance of the matter under discussion. The apostle Paul wrote that people are “without excuse” regarding their failure to acknowledge God properly, because what we can know of God is “manifest” in us. Paul’s assertion is that “the things that are made” deliver such evidence. If that is true, evidence for God is virtually all around us and failure to recognize God through nature a serious issue (as indeed Paul goes on to argue). If it is false, atheists can be a lot more relaxed.

But the above verses from Romans serve us also to delineate the interesting epistemic ways how to detect design.

Apparently, Paul construes nature as evidence that purportedly enables people to “see” God’s invisible attributes. That sounds much like a perceptual claim. However – and that’s why I chose the King James Version here

We may therefore construe Romans 1:19-20 as making the following two claims[4]:

(1) Design in nature can be detected in a way analogous to sense perception.

(2) Design in nature can be understood via an inference that the average person can make[5].

Let’s evaluate those two claims in turn.

2.   Evaluating different reconstructions of design detection

In what follows, I shall examine three approaches to the question how we detect design: (1) the Reidian Perceptual Model (RPM) and two inferential approaches, namely (2) Inference to the Best Explanation (IBE) and (3) Bayes’s theorem (BT). William Dembski’s Eliminative Design Inference, though a major contribution in this field, is intentionally ignored. While his design-detection criterion of complexity seems intuitively graspable, the other two (contingency and specification) require some advanced knowledge of the laws of nature and classical statistics. That prevents its being available to everybody and hence makes it unsuitable as a reconstruction of design detection in the Pauline sense.

2.1  The perceptual model (RPM)

Perhaps design is something that we just perceive without making any inferences to it. A glimpse of this view can be found in the famous passage of William Paley’s Natural Theology:

…we perceive (what we could not discover in the stone) that its several parts are framed and put together for a purpose.[6]

Alvin Plantinga reflects this view:

I encounter something that looks designed and form the belief that it is designed: perhaps this isn’t a matter of argument at all (anymore than in the case of perception or other minds)

The key idea is that we detect design in a way analogous to sense perception and the perception of other minds. Let me therefore quickly present the underlying Reidian theory of perception, which I shall henceforth call RPM (Reidian Perceptual Model):

When I grasp an ivory ball in my hand, I feel a certain sensation of touch. In the sensation there is nothing external, nothing corporeal. The sensation is neither round nor hard; it is an act of feeling of the mind, from which I cannot, by reasoning, infer the existence of any body. But, by the constitution of my nature, the sensation carries along with it the conception and belief of a round hard body really existing in my hand.[8]

Reid holds that our perceptual beliefs are non-inferential, but triggered by our sensory experiences. More specifically, resisting to have the perceptual belief in question is

…not in my power…My belief is carried along by perception, as irresistibly as my body by the earth.[9]

The same goes for the perception of other people as living creatures.

As to design, Reid assumes that it is primarily a mental property and only secondarily a material one[10]. On that presupposition, design detection cannot be based on past experiences. We may have seen a potter form a vase and therefore correctly infer a potter-designer when we find a vase out in the forest. But we have never seen the mental quality “design” in the potter’s mind. Furthermore, Reid holds, even in our own case we often do not know mental qualities directly, but through signs and effects. Taking a relevant example, I may become aware of my pottery qualities only when I exercise them.

In summary, according to the Reidian theory, we detect certain qualities with certainty through our innate mental faculties: physical qualities through our sense perception; mental qualities of other people, including design, through our “design-detecting faculty”. Of course, all those require signs and effects whose occurrence is being perceived through the common gateway, viz. sense perception.

It seems reasonable prima facie to believe that design is being detected in this way. In everyday life, we do not hesitate to ascribe design when we find a watch in the meadow, see a heart written into the sand, or encounter a fireplace deep in the forest. Still, there are objections to it.

Objection 1: Missing account of design marks

As said above, the presented model hinges on the detection of signs or effects which unavoidably lead to the formation of a perceptual or design belief. It should therefore be possible to give an account of those signs in the case of design. Reid himself loosely mentions some criteria like contrivance, intent or beauty, but apparently does not try to delineate them sharply. Contemporary accounts like Dembski’s Eliminative Design Inference or Behe’s Irreducible Complexity cannot be integrated in the RPM because they are highly inferential. Complexity alone won’t do: when I look at my laptop’s case, it is precisely its sleek and minimalistic (hence uncomplex) look that reinforces my belief that it was designed. Purpose alone won’t do either: if I found a strange apparatus in a cave, I wouldn’t doubt its human origin even if I had no clue what its purpose is. Maybe a combination of complexity and purpose is the key, but that needs to be elaborated.

Do we need to find an exhaustive definition at all? Even Dembski concedes that at its very bottom, detecting design is something we cannot rationally reconstruct:

[I]dentifying suitable patterns and side information for eliminating chance requires of [the inferer] S insight…. What’s needed is insight, and insight admits no hard and fast rules…. We have such insights all the time, and use them to identify patterns and side information that eliminate chance. But the logic of discovery by which we identify such patterns and side information is largely a mystery.[11]

Claiming a mystery, however, is an unsatisfying philosophical move; rather, it should be attempted to give a comprehensive account of design marks that are instantly available to everybody[12]. As long as this gap persists, the objection should be regarded as a mild undercutting defeater[13].

Objection 2: Cases too complicated for design perception

Timothy McGrew objects that there are cases too complicated for a quick (perceptual) design detection[14]. He concedes that the “analogy with visual perception is very attractive” and that it “is a poor handyman who needs the help of philosophy to distinguish the lawnmower from the lawn”[15]. He then gives some examples in which he thinks a design detection is not readily, i.e. perceptually made. I will now consider whether those examples pertain to the design argument, and even if they do, whether a satisfying reply from the perceptual standpoint can be given.

The first example is Spenser’s Amoretti. McGrew asks if Spenser deliberately ordered the sonnets so as to reflect the liturgical year or if their spacing is a coincidence. Indeed, this is not easy to judge. But I doubt that it is relevant for the purposes of the design argument. As McGrew admits, Spenser undoubtedly “designed” the Amoretti. The design argument hinges on the question whether living things have been designed as a whole, not if the order and relations within the things was so intended to the last minute detail. For example, we need not decide if the flexibility of the elephant’s proboscis compensates for the rigidity of his neck[16] (and so is designed in this respect) or if the two features just coexist. The only crucial thing is to show that the elephant as a whole is a product of design.

Second, McGrew puts forward the example of the fossil of a feathered dinosaur and asks whether the feathers are part of the dinosaur or the product of some hoax. This example, however, is question-begging. The latter explanation undoubtedly contains design and the former possibly does so. McGrew, however, seems to presuppose that the feathers being an integral part of the animal means that the whole animal is a product of chance and/or natural law. But this is to use as the conclusion as a premise.

His third example is a criminalistic one: “Shall we classify the untimely demise of Birdy Edwards as a homicide or as an accidental death?”[17] I think this example, too, misses the point, because it is about an event. The design argument deals with things made of matter, not with the reconstruction of past events. So if we confine the evidence relevant for the design argument to material things, this example becomes irrelevant.

Hitherto, McGrew has not shown that there are counterexamples disabling a perceptual view of design detection. But we can go even further and show that the proponent of the perceptual model has a good answer to cases in which it seems impossible to perceive design. Take the prime number pattern from the movie Contact. The sequence of numbers is per se certainly not identifiable as design. But once its character as a prime number string is identified, it is instantly perceived as design. So the Reidian does not need to insist that all instances of design are being perceived at once (for indeed many aren’t), only that the basic marks of design are. It may take some pains to discover them, but once they’re discovered, no more inference is needed to establish a detection of design.

However, if we grant this, an extreme perceptual view of design beliefs as “basic”[18] becomes implausible. Remember that Plantinga claims

…the belief that something or other is a product of design is not formed by way of inference, but in the basic way…[19]

Precluding inferences completely does not work in the Contact case nor in many other relevant settings. As objections 3 and 4 will show, the RPM needs to be complemented by an inferential element.

Objection 3: Disagreements about the occurrence of design

What if people strongly disagree whether a certain phenomenon exhibits design or not, as is most notably the case with living things?

Indeed, people so design-hostile as Charles Darwin, Richard Dawkins and Francis Crick admitted that the appearance of design pressed hard on them:

[I]n the course of that conversation I said to Mr. Darwin, with reference to some of his own remarkable works on the Fertilisation of Orchids, and upon The Earthworms, and various other observations he made of the wonderful contrivances for certain purposes in nature—I said it was impossible to look at these without seeing that they were the effect and the expression of Mind. I shall never forget Mr. Darwin’s answer. He looked at me very hard and said, “Well, that often comes over me with overwhelming force; but at other times,” and he shook his head vaguely, adding, “it seems to go away.”[20]

Biologists must constantly keep in mind that what they see was not designed, but rather evolved.[21]

Biology is the study of complicated things that have the appearance of having been designed with a purpose.[22]

The Reidian model fits neatly with those accounts, especially the aspect of the design impression “forcing itself” on one. It seems plausible to explain the strong disagreements with philosophical preconceptions. But this explanation is nothing RPM can yield. Why is design detection in some cases so vulnerable to prejudices, while sense perception, to which it is analogous according to Reid, seems resilient to such usurpations? As we will see, inferences are the answer to this riddle.

Objection 4: Dissimilarity to sense perception

It was insinuated in objection 3 but now becomes obvious: beliefs formed by sense perception and beliefs formed by design perception are more dissimilar than the RPM suggests. Reid claims that we perceive design in quite the same way in which we perceive a stone in our hand. But this is not so.

When I touch a stone, my touch-sensitive nerves are instantly triggered and through neuronal pathways cause a perceptual mental experience in me. There is no intermediate between the object (the stone) and my sense experience than just those biological pathways. In design perception, it also is a perceptual mental experience that the thing at hand causes. Now Reid (and as it seems, also Plantinga) argues that this perceptual experience is of such a kind that it immediately, en passant, causes a design belief. But if this were true, it would be as hard to disbelieve design as it is to disbelieve the tactile perception of a stone. (Of course I can disbelieve that the thing in my hand is a stone, but I cannot disbelieve that something is in my hand, unless I’m a radical skeptic of the Matrix kind). This seems highly implausible: There is, for instance, no reason to doubt that Richard Dawkins does believe that living things are the products of blind natural processes.

It seems that design beliefs are much more vulnerable than ordinary perceptual beliefs. A good explanation for that is that design beliefs contain an inferential element. Daniel von Wachter reflects this by arguing that perceptual beliefs and inferential beliefs are criticized in different ways:

Miller’s perceptual belief that there is a cow because he has seen it can be criticised by suggesting that his eyes do not work properly: by pointing out that there is an elk which one can easily mistake for a cow, or that he is under the influence of drugs, or that there is an evil demon who manipulates his mind. Jones’ inferential belief that there was a cow in his garden because he has seen the hoof marks can be criticised by suggesting a better explanation for the data: by pointing out that an elk’s hoof marks look similar, or that Smith had said that he intended to deceive Jones by producing marks in the garden.[23]

Now, Reid’s intriguing picture of an innate “design-detector” seems in danger. But we need not abandon it. I claim that this mental capacity consists in having a mental picture of how a designer works. We have this mental picture because we are ourselves “designers”. By knowing which reasons and methods a designer might have to contrive something, we detect design so easily – in most cases. So it is because we have learned (inferentially!) from ourselves and other human beings how designers work that we detect design so routinely. This account preserves the Reidian “congenitality” of the design-detector while including inferences.

In summary, it seems we need to include inferences to explain our mental “design-detector” as well as to explain uncertainties or disagreements in some putative design cases. Let’s therefore explore two models of inference-making.

2.2  Inference to the Best Explanation

An inference to the best explanation (IBE) rests upon an act of “abduction” that C.S. Peirce describes as follows:

The surprising fact C is observed;

But if A were true, C would be a matter of course,

Hence, there is reason to suspect that A is true.[24]

Now, this first step of abduction can merely generate some hypotheses A1…An which seem plausible. Which one is judged best will depend on other factors such as simplicity, explanatory power, analogy, aesthetics, and detailedness of explanation[25].

One advantage of IBE is that it comes in degrees. In contrast to Dembski’s Eliminative Design Inference, the design hypothesis need not be so mighty as to rule out all other possibilities. It just needs to be the most plausible one. This contrast also pertains to a comparison between IBE and RPM, for RPM consists in a “digital” detection of design. Second, IBE involves considerations of causal adequacy. To use an example of McGrew: left with the choice to ascribe the Winged Victory of Samothrace to a desert sandstorm or a newborn infant we would wonder if a third alternative hasn’t been overlooked.
All in all, IBE seems to go well with our intuitions concerning design. Could it be that putative perceptual design detections are really inferences to the best explanation at hyper-speed, as Sherlock Holmes describes it:

From long habit the train of thoughts ran so swiftly through my mind that I arrived at the conclusion without being conscious of intermediate steps.[26]

This might be the case if we construe a design detection as follows:

  1. We know by experience from ourselves and other people what designers can do and what reasons and methods they employ.
  2. We detect marks of design directly in a way we recognize other people.
  3. The detection of design is basically an inference.
  4. When things strongly exhibit design, the design hypothesis is unhesitatingly chosen as the true explanation because of 1. and 2.
  5. In other, less clear cases, competing hypotheses are also considered. A design detection can then occur as a “false positive” or be missing as a “false negative”.

Still, two serious objections against IBE have been levelled. First, the initial abductive step may falsely ignore the true hypothesis; we then end up choosing the “best of a bad lot”. An IBE proponent may reply here that it is a matter of intellectual virtue of the individual examiner to avoid exclusion of hypotheses out of negligence or aversion. But surely, no one can consider infinitely many hypotheses; we need to restrict ourselves, and the crucial question is how. IBE gives no criteria for sorting out hypotheses.

Second, some of the criteria for ranking rival hypotheses are debatable, especially the criterion of aesthetics. The beauty of an explanation does not necessarily entail its truth, nor does its simplicity. The fact that some scientists like Lavoisier or Darwin used simplicity and analogy as methodological guides to adjudicate between hypotheses is not a good argument either, for some aspects of their theories are simply untrue. McGrew sums this problem up as follows:

[I]f IBE is to stand on its own as a methodology, then we need an account of the link between lovely and likely theories that does more than appeal to the practice of scientists or the aesthetic attractions of notions like simplicity.[27]

A last disadvantage of IBE as a model for design detection might arise from its being an inferential model. Alvin Plantinga points out that none of the criticisms applying to inferentially formed beliefs pertain to perceptually formed beliefs[28]. If that is true, and if immunity against criticism indeed raises the justification of a belief, then perceptually formed beliefs are generally more reliable than inferentially formed ones. However, as I pointed out in section 2.1 (Objection 4), the RPM should not be construed as a strong analogy to sense perception; therefore, from this objection against IBE does not follow too much profit for RPM.

The upshot is that IBE has great initial plausibility and might even be merged with RPM into a powerful model that explains design detections (or the absence thereof) in all kinds of cases. It still has serious problems regarding the choice of hypotheses and the adjudication between hypotheses. There is, however, a model which mends the weaknesses of IBE.

2.3  Bayes’s theorem

Suppose you encounter some phenomenon (which in a moment we shall call evidence). You form some hypotheses as to how the phenomenon can be explained. Your concern is to find the best explanation for the phenomenon. Alternatively, you have some hypotheses and seek to identify the one that best fits the evidential data.

Bayes’s theorem is a simple yet powerful mathematical tool that helps you in both cases. It consists of the term

with h being the hypothesis, e being the evidence and k our background knowledge.


  • P(h/e&k) represents the posterior probability of h (“probability of h given e and k”)
  • P(e/h&k) is also called the likelihood of e (“probability of e given h and k”)
  • P(h/k) is the prior probability of h (“probability of h being true given k”)
  • P(e/k) is the prior probability or expectedness of e (“probability of e given k”)

As k figures in all sub-terms, I will omit it henceforth, at least formally. It continues to play a natural role in our discussion of competing hypotheses (see section 3).

Bayes’s theorem has three advantages over IBE:

i) It gives all logically possible hypotheses a “fair chance”, adjudicating between them via a quantified comparison of probabilites (not possible with IBE).

ii) It captures our intuition that a hypothesis’s low prior probability should lower its posterior probability.

iii) It accounts for the predictive power of a hypothesis: if an event with low expectedness is being predicted by it, that raises the posterior probability of the hypothesis.

The points ii) and iii) help to quantify the quality of a hypothesis, something not possible with IBE, which must rely on subjective assessments.

However, one problem of Bayesianism arises precisely out of these two points. How to assign numerical values to propositions such as “The prior probability of God’s existence is…” or “The expectedness of there being a universe is…”? Subjective Bayesians hold that one can just pick subjective values, which is a violation of objectivity and might leave people far apart in their final judgment about a hypothesis.

A better approach is Realistic Bayesianism[29], which starts with a precise determination of the probabilistic interrelationships of available hypotheses and then inserts intervals of values to eschew the unrealistic expectation of determining exact values.

What if even such fuzzy value assignments seem impossible? We need not worry about that as soon as we have a partition, that is an exhaustive set of mutually excluding hypotheses (whose probabilities add up to one). We can then compare their posterior probabilities without assigning concrete numbers. True, the assignment of prior probabilities needed to compare posterior probabilities remains irremediably debatable; but as the pertinent criteria such as simplicity etc. are here being applied to sub-terms of a larger calculation, we can expect more accurate judgments. Another advantage is that the denominator  can be omitted, for it figures in the calculations of all hypotheses.

Thus construed, BT is an attractive approach for comparing the probabilities of different explanations. How does it fare compared to the perceptual model? Timothy McGrew puts forward seven criteria for assessing the quality of a rational reconstruction of the design argument[30]:

  1. Integration into a larger epistemic framework
  2. Topic Neutrality
  3. Objectivity
  4. Sensitivity to relevant parameters
  5. Relevance of captured parameters
  6. Accessibility of information needed for the reconstruction
  7. Explanatory Power especially concerning disagreements

On criterion I, the RPM seems to do better. Reid’s theory of design as a mental quality corresponds to many day-by-day experiences we make. Inferences seem to be the domain of criminologists and scientists (notwithstanding the Reidian reply that marks of design are perceived instantly once they’re discovered). The RPM becomes even stronger in a theistic framework. If God put the mental capacity to discern design from not-design in us (as Rom 1:19-20 could be read), this reinforces the integration of the model. BT is rooted in probability calculus, which also represents an epistemic framework, however less intuitive.

As far as topic neutrality is concerned, BT clearly seems to win the day. Design perception is applicable to design cases only; BT can be used for scientific theories, for discerning between motives and much more.

Objectivity also speaks in favor of BT, though only slightly. The area of interest here, namely design in living things, is ridden with disagreement. An externalization of such disagreement seems impossible with RPM; either people see design, or they don’t, no debate possible. Here it must surrender to BT, which still suffers from irremediable subjectivity, but at least the mathematical framework is agreed upon.

As to sensitivity, both seem to be sufficiently sensitive. If design is indeed a mental quality, it is readily perceived whenever at hand; BT accounts for parameters like the prior probabilities and thus accommodates for example for the awkwardness of some hypothesis or the unexpectedness of a phenomenon. Both track down relevant parameters as well, for the very same reasons.

Accessibility seems to be the domain of the RPM, but only at first glance. True, perceptual information is most readily accessible; but what about those that see the animal and still have no impression of design? Nothing can be mended there, because there are no premises that could be modified. BT offers moderate accessibility with the possibility of discussing the criteria picked for example to assess the prior probability of a hypothesis.

Lastly, the explanatory power of BT seems greater than that of RPM. As said before, RPM cannot by itself explain the disagreements concerning design detection. Disagreements on a BT account, conversely, can be tracked to disagreements in the assessment of the likelihood and/or prior probability. There seems to be more room for rational dispute than with PM.

Apparently, each model has its advantages and drawbacks. Therefore, as insinuated in section 2.2, I will now attempt a synthesis of RPM and BT.

2.4  The BIBOP model: a new synthesis

BIBOP stands for “Bayesian inference based on perception”. It assumes that design detections are not perceptions simpliciter, but Bayesian inferences based on the perception of design marks. It resembles the model proposed as a fusion of design perception and IBE, now with the advantages of BT:

  1. We know by experience from ourselves and other people what designers can do and what reasons and methods they employ. We also know roughly the scope of natural processes.
  2. We detect marks of design directly in a way we recognize other people.
  3. The detection of design (or non-design, for that matter) is basically an inference.
  4. When we see a thing T and think about its origin, three classes of hypotheses are available in our minds: HD (design), HC (chance) and HN (necessity through natural law). These hypotheses constitute a partition.
  5. When things strongly exhibit design, we instantly assign P(T/HC) = 0 and P(T/HN) = 0 because of 1., thus leading to P(HC/T) = 0 and P(HN/T) = 0. Necessarily then, P(T/HD) = 1 because of 4. (and buttressed by 2.). This explains why the impression of “design perception” is so strong to us.
  6. In other, less clear cases, or in cases where for other reasons HC and HN get much credit, P(T/HC) > 0 and/or P(T/HN) > 0. In those cases, the underlying inferential character of the belief formed becomes obvious. A design detection can then occur as a “false positive” or be missing as a “false negative”, which powerfully explains disagreements.

Let’s now test whether in a concrete case – the question of design in living things – the BIBOP model can indeed prove itself.

3.   The BIBOP model put to the test

The hypotheses I will consider as explanations for the existence of living things are the following:

  • HT: The hypothesis of theism. This replaces the generic design hypothesis. HT includes all the aspects of HD, but goes further, thus lowering its prior probability but increasing its explanatory power. In total, I believe HT to be more powerful. Anyway, it is the ultimate goal of any design argument.
  • HE: The hypothesis of Darwinian/synthetic evolution. It replaces the generic chance hypothesis from above. The synthetic evolutionary theory heavily relies on chance (random genetic mutations) but includes natural law to such a degree that it is a far more plausible hypothesis than raw chance.
  • HN: The hypothesis of natural law. This idea has been proposed mainly by Michael Denton[31]. He claims that the evolution of life, even of human beings is “encoded” in the laws of nature and hence a matter of necessity once the natural laws are implemented in the universe.

It is important to note that these hypotheses are mutually exclusive and that they constitute a partition, i.e. their probabilities add up to one.

Now let’s check the BIBOP model with those hypotheses:

1’. (taken for granted)

2’. Do we indeed detect marks of design in living things? I can’t see any reason to deny this. Remember, even fierce proponents of HE admit that they can hardly resist the impression of design in living things; surely this is due to some marks they see. Although it is beyond the scope of this essay to give a full account of design marks, we can name a few that occur in living things and are recognizable without a MSc in molecular biology: complexity; purposiveness; playfulness; beauty.

3’. (taken for granted)

4’. In our present examination, HT, HE and HN (though in most people’s minds only HT and HE) are available.

5’. Considering the vast amount of beauty, complexity and purposiveness in nature, a person with HE or HN will naturally assign P (T/HT) » 1. Hence, P (T/HE) and P (T/HN) » 0. Such a person “cannot help” but believe that living things were designed by a very powerful, good God.

6’. If P(HE) or P(HN) > 0, all depends just on how great this probability is. At any rate, we can explain strong disagreements at this level. If someone fiercely believes in Darwinian evolution out of an atheistic bias, he will a priori assign P (T/HE) » 1.

So far, the BIBOP model has done a good job in explaining how people come to hold their beliefs in designed (or evolved) living things. Of course, more hypotheses than the three proposed ones can be cited. The main thing is that they form a partition.

Let me still anticipatively reply to an objection that may be raised against 5’. Does not the cruelty and mercilessness we so often seem to observe in the animal kingdom speak against a good God as designer? This objection is substantial, but not unanswerable. For one, the beauty of nature is overwhelming and in my view far outweighs the seemingly cruel aspects. Second, what are the alternatives to a good God? An evil God seems absurd, considering all the beauty. A better option seems to be a good God which is less than all-powerful and has partially “lost control” over his creation. But that still calls out for an explanation where the evil bits come from. I believe that such an explanation can be given, even without dispensing with God’s omnipotence; it is not the aim of this essay to do this. Another possibility is the one chosen by the ID movement: suspending judgment on the designer’s identity. But that leaves us with an unnecessarily tentative hypothesis. It misses the chance to explain some observed features of living things with the designer’s motives and is open to the criticism of Elliott Sober who justly complains that ID makes no independently testable predictions as to which kinds of living things the designer would make[32]. All I would like to emphasize here is that theological objections against the theistic hypothesis can in principle be deflected.

In summary, the advantages of BIBOP remain:

  • no complicated inferential arguments for how to detect design marks needed
  • applicable to clear as well as opaque cases
  • accounts for disagreements
  • shifts the onus away from the design proponent and thus confers justification to the belief that living things have been designed, assigning to HE the strength of an undercutting defeater instead of that of a rebutting defeater

4.   Prospects and Perspectives

As said before, BIBOP helps maintain that any theist is rational in holding his belief in divine design of living things. If one wishes to gut further and show that it is irrational to hold ØHT, one needs to do two things:

  • show that P (T/HE/N) is indeed very low
  • show that P (T/HT) is high (compared to P (T/HE) but also to P (T/HD)).

Both can be done, and, as I believe, those projects will turn out to be nothing else than the anatomizing of intuitions we all have: that complex living things don’t arise either by chance or natural law and that the theistic God has good reasons to make a world like ours. If these arguments succeed, we will have gone full circle from our starting point: people are “without excuse” in denying God’s designerhood in the face of the evidence, because the double power of perception and inference speak in favor of it.

[1] Romans 1:19-20 (KJV)

[2] (O’Connor 2003), p. 65

[3] The Nestle-Aland Greek New Testament (see, visited on Nov 1, 2017) renders Rom 1:20 with both the verbs kathorao (meaning to behold fully, to clearly see) and noieo (meaning to exercise the mind, to comprehend). While the KJV reflects that, the ESV speaks of “clear perception”, which omits the inferential nature of noieo.

[4] I am aware that Rom 1:19-20 was not intended as a design argument. Paul’s aim is to show that all people are rightfully under God’s wrath because their failure to worship God properly is inexcusable. Still, he makes a connection between nature and God as Creator. This connection might be construed in one of the following two ways: (1) People know of God independent of and prior to their perception of nature – maybe through a sensus divinitatis – with the observation of nature providing strong support to their “God-sense” (2) People know of God primarily through the observation of nature. Either way, nature provides at least strong support for the belief that the theistic God exists.

[5] I added „the average person“ because some design inferences may be too mathematical to be doable “on the spot“. William Dembski’s Eliminative Design Inference (Dembski 1998) is an example.

[6] (Paley 1809), pp. 1–2 [Emphasis added]

[7] (Plantinga 2011) (Kindle Locations 3570-3571).

[8] (Reid 1872a), p. 450

[9] (Reid 1872b), p.183-184

[10] (Reid 1872a), p. 503

[11] (Dembski 1998), p. 148

[12] I consider it as important to find such a universally accepted account. The judgment whether living things are the products of design or not should be open for everyone, not just scientists.

[13] A term I borrow from (Plantinga 2011) (Kindle Locations 2410-2411).

[14] (McGrew 2005), p. 253

[15] ibd.

[16] One of Paley’s less convincing “design” examples.

[17] ibd.

[18] As (Plantinga 2011) puts it. Von Wachter reads Plantinga as being neutral about whether a basic belief is supported by perception or not. (Von Wachter 2014), p. 59

[19] (Plantinga 2011) (Kindle Location 3572)

[20] (Argyll 1885), cited in: (Ratzsch 2003), p. 124

[21] (Crick 1988),
p 138

[22] (Dawkins 1986), p. 1

[23] (Von Wachter 2014), p. 60

[24] (Peirce 1935), p.189. (Cited in (McGrew 2005), p. 273)

[25] The lists and terms differ among authors such as Swinburne, Thagard or Lipton.

[26] (Doyle 1905)

[27] (McGrew 2005). p. 282

[28] (Plantinga 2011) (Kindle Locations 3655-3659)

[29] (McGrew 2005), p. 288

[30](McGrew 2005), p. 256-61

[31] (Denton 1998)

[32](Sober 2003), p. 36



Argyll, George Douglas Campbell, 8th Duke of. 1885. “What Is Science?” Good Words, no. April:236–45.

Crick, Francis. 1988. What Mad Pursuit. London: Penguin.

Dawkins, Richard. 1986. The Blind Watchmaker: Why the Evidence of Evolution Reveals a Universe without Design by Unknown. Unknown edition. W. W. Norton & Company.

Dembski, William A. 1998. The Design Inference: Eliminating Chance through Small Probabilities. New. Cambridge ; New York: Cambridge University Press.

Denton, Michael. 1998. Nature’s Destiny: How the Laws of Biology Reveal Purpose in the Universe. Original ed. edition. New York: Free Press.

Doyle, A.C. 1905. “A Study in Scarlet.” In The Complete Sherlock Holmes. Garden City, New Jersey: Doubleday.

McGrew, Timothy. 2005. “Toward a Rational Reconstruction of Design Inferences.” Philosophia Christi 7 (2):253–98.

O’Connor, Robert. 2003. “The Design Inference. Old Wine in New Wineskins.” In God and Design. The Teleological Argument and Modern Science. Routledge.

Paley, William. 1809. Natural Theology: Or, Evidences of the Existence and Attributes of the Deity. 12th ed. J. Faulder.

Peirce, C.S. 1935. Collected Papers. Edited by C. Hartshorne and P. Weiss. Cambridge, MA: Harvard University Press.

Plantinga, Alvin. 2011. Where the Conflict Really Lies: Science, Religion, and Naturalism. 1st ed. Oxford University Press.

Ratzsch, Del. 2003. “Perceiving Design.” In God and Design. The Teleological Argument and Modern Science. Routledge.

Reid, Thomas. 1872a. “Essays on the Intellectual Powers of Man.” In The Works of Thomas Reid, 7th ed. Vol. I. Edinburgh: Maclachlan & Stewart.

———. 1872b. “Inquiry into the Human Mind.” In The Works of Thomas Reid, 7th ed. Vol. I. Edinburgh: Maclachlan & Stewart.

Sober, Elliott. 2003. “The Design Argument.” In God and Design. The Telelogical Argument and Modern Science. Routledge.

Von Wachter, Daniel. 2014. “Defending Design Arguments Against Plantinga.” Philosophia Reformata 79:54–65.

No Comments

Post A Comment